关键词: 
小波阈值 /
平滑去噪 /
能谱 /
Matlab 程序

摘要: 采用小波软、硬阈值函数折中后的小波阈值方法对天然放射性环境下 谱数据进行平滑处理。通过Matlab数据分析软件，分析对比了五种不同的平滑去噪方法：五点平滑法、傅里叶变换法、小波软阈值法、小波硬阈值法，以及折中的小波阈值方法。结果表明：折中的小波阈值方法在反映 能谱特征的峰本比和峰总比分别为7.160 9，0.394 5，平滑去噪效果要明显优于其他四种方法，能最大程度地保留原始信号特征，具有较高的应用价值。This work illustrated a wavelet threshold method by combining soft and hard threshold function of wavelet for energy spectrum under the natural radioactive condition as a kind of smooth processing. By using Matlab Program, we compared five different smooth de-noising methods including smoothing by five points, the Fourier transform method, soft and hard threshold method and average wavelet threshold method.We showed that the average wavelet threshold method can reflect the ratio of the peak and the peak-to-total ratio of characteristic peak of energy spectrum as 7.1609 and 0.394 5, respectively. Thus, this method is superior to the other four methods in de-nosing, and it is able to keep the original signal properties with the largest extent and has a high value of application.